BACTERIA IN WATER. 631 
by colored lines, and the number of colonies in several squares is 
counted in order to obtain an average and estimate the entire 
number, 
Water which contains numerous liquefying bacteria had better 
be examined by the use of nutrient agar instead of gelatin; and in 
very warm weather it will be necessary to use an agar medium, as 
ten-per-cent gelatin is likely to melt if the temperature goes above 
22°C. A difficulty in the use of agar for plates consists in the lia- 
bility of the film to slip from the glass. This may be remedied to 
some extent by adding a few drops of a concentrated solution of gum 
acacia to the liquefied agar medium. Petri’s dishes are well adapted 
for the use of the agar medium, as the objection referred to does not 
apply to them. The gelatin-agar medium, containing 5 per cent 
of gelatin and 0.75 per cent of agar, may also be used with advan- 
tage in the bacteriological analysis of water. Much stress was at 
one time laid upon the enumeration of liquefying colonies, upon 
the supposition that the liquefying bacteria were especially harmful 
as compared with the non-liquefying, and that a water containing 
many liquefying colonies was to be looked upon with suspicion. We 
now know, however, that there are many common and harmless 
saprophytes which cause the liquefaction of gelatin, and that some 
of the most dangerous pathogenic bacteria do not liquefy gelatin. 
This distinction has therefore no special value, and the question for 
bacteriologists to-day is not how large is the comparative number of 
liquefying colonies, but what species are represented by the colonies 
present, liquafying and non-liquefying, and what are the special 
pathogenic properties of each. The answer to these questions, in 
the case of any particular water supply, calls for special knowledge 
and great patience and care in the isolation in pure cultures, and 
careful study of the various species present. 
It is now generally recognized that a mere enumeration of the 
number of colonies which develop from a water under investigation 
is not a sufficient indication upon which to found an opinion as to its 
potability. An excessive number of bacteria is an indication that 
the water contains a large amount of the organic material which 
serves as pabulum for these microédrganisms. But the chemists are 
able to determine the amount of organic matter present in water 
with greater precision ; and, as we have seen, the number of bacteria 
may increase many-fold in water which is kept standing in the labo- 
ratory for two or three days in a well-corked bottle. As a matter of 
fact, the enumeration of bacteria in water, although it has given us 
results of scientific interest, has not materially added to the methods 
previously applied for estimating the sanitary value of water ob- 
tained from various sources for drinking purposes. But the bacte- 
